Assembly and kinetic properties of myosin light chain isozymes from fast skeletal muscle.

Myosin from chicken pectoralis muscle consists of isozymes that differ in their alkali light chains. It is possible to isolate alkali 1 (A1) and alkali 2 (A2) homodimers of native myosin by immunoadsorption methods, and to compare their steady-state kinetics as well as their assembly into synthetic filaments under a variety of ionic conditions. Bipolar filaments of the isozymes formed at low salt concentrations had a narrow length distribution and did not differ from controls made from unfractionated myosin. Chicken myosin also assembles into highly homogeneous minifilaments similar to those formed by rabbit myosin in a citrate/Tris buffer. Analytical ultracentrifugation and electron microscopy showed that A1-homodimer, A2-homodimer and unfractionated myosin assembled into 0.3 micron short, bipolar minifilaments, which were indistinguishable from one another in size and shape. The steady-state myosin ATPase activity of the two homodimeric isozymes was identical in K+(EDTA) and Ca2+ assay media. The actomyosin Mg2+ ATPase measured at 25 and 55 mM-KCl (pH 8.0) showed only minor differences in both Vmax and Kapp. Actomyosin activity was also determined for the more homogeneous minifilament preparations of the isozymes and these, as well, produced essentially indistinguishable kinetic parameters. Thus we find no evidence to support the hypothesis that a particular alkali light chain of myosin can affect either the structure of the filaments or the steady-state rate of ATP hydrolysis.

Disassembly of synthetic 10-nm desmin filaments from smooth muscle into protofilaments.

Synthetic 10-nm filaments formed from highly purified turkey gizzard desmin have a helically oriented substructure and disassemble into 2 to 2.5 nm protofilaments and 3.5 to 5 nm subfilaments after treatment with 1 or 2 M urea or with low-ionic-strength buffer (8 mM Tris, pH 8.2). SDS-gels of 10-nm filaments treated with these solutions show that a single 55 000-dalton band is present before and after all treatments and indicate that the newly revealed substructure is not caused by proteolysis. Antibodies to electrophoretically purified desmin react, in immunodiffusion, only with the antigen and decorate both the subfilamentous particles and the synthetic 10-nm filaments. These studies indicate that a synthetic 10-nm desmin filament is a rope-like structure constructed from the 2 to 2.5 nm diameter protofilaments.

Immunoelectron and immunofluorescence localization of desmin in mature avian muscles.

Antisera or affinity-purified antibodies shown to be specific for avian gizzard desmin antigen by double immunodiffusion, antigen-blocking, and immunoautoradiography experiments have been used in indirect immunofluorescence and indirect immunoelectron microscopy to demonstrate localization of desmin in myofibrils from mature avian muscles. The light microscope results agree with the work of others in that they suggest that desmin is primarily at or near the periphery of Z-lines of striated muscle myofibrils. Immunoperoxidase labelling more clearly shows that the reactive desmin antigen is located almost entirely between Z-lines of adjacent parallel myofibrils and that there is no obvious correlation between locations of T-tubules and desmin structures. The electron-dense reaction product often followed an approximately linear course between Z-lines of adjacent myofibrils and indicated the desmin antibodies had decorated a small number of filaments spanning this region. These results suggest that the desmin found in close association with myofibrils of mature striated muscle is in the aggregated form of 10-nm filaments.

Integrin expression in developing smooth muscle cells.

We studied the specific expression patterns and distributions of alpha1 and beta1 integrin subunits, the major cell adhesion receptors in smooth muscle, in developing smooth muscle cells from 16-, 18-, and 20-day embryonic gizzards and from 1- and 7-day post hatch chick gizzards by SDS-PAGE, immunoblotting, and immunoelectron microscopy. Antibodies raised against alpha1 and beta1 integrins isolated from avian gizzards were used as probes. Gels and blots showed that the amount of alpha1 and beta1 integrins increased as age increased, with major increases at 1 and 7 days post hatch. Image analysis of immunoelectron micrographs demonstrated that statistically significant labeling increases occurred between embryonic Days 16 and 18, between embryonic Day 20 and 1 day post hatch, and between 1 day and 7 days post hatch. Immunolabeling with both anti-alpha1 and anti-beta1 integrin was prominent at membrane-associated dense plaques (MADPs) and at filament anchoring regions at cell ends. This indicates that alpha1 and beta1 integrin expression coincides temporally with the intracellular proliferation and reorientation of myofilaments. The similarity in distribution patterns of alpha1 and beta1 integrins during development suggests that the two integrin subunits are synchronously expressed during development and do not appear sequentially. (J Histochem Cytochem 46:119-125, 1998)

Regulatory role of arginine-specific mono(ADP-ribosyl)transferase in muscle cells.

Earlier we demonstrated that meta-iodobenzylguanidine (MIBG), a specific inhibitor of arginine mono-ADP-ribosylation blocks proliferation and differentiation of chick skeletal myogenic cells in culture (Exp. Cell Res., 1992, 201:33-42). Membrane fractions from 4-day, myotube cultures of embryonic chick muscle cells were incubated with 32P-NAD+. Several proteins were labeled, but labeling of two hands of about 53 and 36 kDa appeared to be due to arginyl ADP-ribosylation. Immunoprecipitation with D3 monoclonal antibody to the intermediate filament protein desmin, SDS-PAGE and autoradiography demonstrated that the 53 kDa band contained desmin, and that this desmin is ADP-ribosylated by the endogenous arginine-specific mono(ADP-ribosyl)transferase (Exp. Cell Res., 1996, in press). Desmin is the muscle-specific intermediate filament protein, and it appears to be one of the first muscle-specific proteins expressed during terminal myogenic differentiation. We have examined whether desmin can be ADP-ribosylated in muscle cells by use of polyclonal antibodies for ADP-ribosylated arginyl residues. We have found that soluble desmin is present in 5-6 day myogenic cell cultures and that this desmin contains ADP-ribose, demonstrating that desmin is ADP-ribosylated in skeletal muscle cells. We also found that purified avian desmin contains antigenic material that reacts with these antibodies. In both cases, NaCl had no effect on the reactivity, but NH2OH did, which is consistent with an arginine-ADPR linkage. In summary, these results suggest that ADP-ribosylation is an important regulatory mechanism in differentiating muscle cells, and that the intermediate filament protein desmin is an important substrate for modification in muscle cells.

Effect of porcine stress syndrome on the solubility and degradation of myofibrillar/cytoskeletal proteins.

This study examined the effect of stress classification (stress-positive, stress-carrier, stress-negative) of pigs on selected properties of postmortem muscle, including protein solubility and degradation of proteins such as titin. Longissimus muscle samples were removed 45 min postslaughter, divided into samples, and stored at 0 to 2 degrees C for analysis at 0, 1, 3, 5, and 7 d postmortem. Whole-muscle samples (homogenates) and purified myofibrils were prepared from each sample for analysis by SDS-PAGE. A portion of each muscle sample also was extracted 1) with a low-ionic-strength solution to obtain a sarcoplasmic protein fraction and 2) with two different high-ionic-strength solutions to obtain a myofibrillar/cytoskeletal protein fraction for measurement of protein solubility and for analysis of extracts by SDS-PAGE. No significant differences were observed between muscle from stress-negative and stress-carrier animals in this study. Sarcoplasmic (P less than .05) and myofibrillar/cytoskeletal (P less than .01) protein solubility was lower in muscle samples from stress-positive animals than in muscle samples from stress-carrier and stress-negative animals at all postmortem times studied. The high molecular weight protein titin was degraded more slowly postmortem in muscle from stress-positive than in muscle from stress-negative animals, as observed by SDS-PAGE analysis of whole-muscle samples (homogenates) an myofibrils. The combination of lowered protein solubility and reduced rate of postmortem degradation of structural proteins such as titin may explain, at least in part, the reduced quality and protein functionality of muscle from stress-positive pigs.

Early postmortem biochemical factors influence tenderness and water-holding capacity of three porcine muscles.

The objective of this study was to determine whether differences in pork tenderness and water-holding capacity could be explained by factors influencing calpain activity and proteolysis. Halothane-negative (HAL-1843 normal) Duroc pigs (n = 16) were slaughtered, and temperature and pH of the longissimus dorsi (LD), semimembranosus (SM), and psoas major (PM) were measured at 30 and 45 min and 1, 6, 12, and 24 h postmortem. Calpastatin activity; mu-calpain activity; and autolysis and proteolysis of titin, nebulin, desmin, and troponin-T were determined on muscle samples from the LD, SM, and PM at early times postmortem. Myofibrils from each muscle were purified to assess myofibril-bound (mu-calpain. Percentage drip loss was determined, and Warner-Bratzler shear (WBS) force was analyzed. Myosin heavy-chain (MHC) isoforms were examined using SDS-PAGE. The pH of PM was lower (P < 0.01) than the pH of LD and SM at 30 and 45 min and 1 h postmortem. The PM had a higher (P < 0.01) percentage of the MHC type IIa/IIx isoforms than the LD. The-LD had the greatest proportion of (P < 0.01) MHC IIb isoforms of any of the muscles. The PM had the lowest (P < 0.01) percentage of MHC IIb isoforms and a greater (P < 0.05) percentage of type I MHC isoforms than the LD and SM. The PM had less (P < 0.01) drip loss after 96 h of storage than the SM and LD. The PM had more desmin degradation (P < 0.01) than the LD and SM at 45 min and 6 h postmortem. Degradation of titin occurred earlier in the PM than the LD and SM. At 45 min postmortem, the PM consistently had some autolysis of mu-calpain, whereas the LD and SM did not. At 6 h postmortem, some autolysis of mu-calpain (80-kDa subunit) was observed in all three muscles. The rapid pH decline and increased rate of autolysis in the PM paralleled an earlier appearance of myofibril-bound mu-calpain. The SM had higher calpastatin activity (P < 0.05) at 45 min, 6 h, and 24 h and had higher WBS values at 48 h (P < 0.01) and 120 h (P < 0.05) postmortem than the LD. At 48 and 120 h postmortem, more degradation of desmin, titin, and nebulin were observed in the LD than in the SM. These results show that mu-calpain activity, mu-calpain autolysis, and protein degradation are associated with differences in pork tenderness and water-holding capacity observed in different muscles.

Control of filament length by the regulatory light chains in skeletal and cardiac myosins.

The possible role of the regulatory light chains (LC2) in in vitro assembly of rabbit skeletal and dog cardiac myosins was examined by formation of minifilaments and synthetic thick filaments. After LC2 was removed, the resulting myosin preparations exhibited little aggregation in 0.5 M KCl and 0.05 M potassium phosphate (pH 6.5). Minifilaments migrated as a single, hypersharp peak during sedimentation velocity, but electron microscopic analysis revealed a more destabilized structure for LC2-deficient minifilaments. Thick filaments were formed in buffers containing 0.15 M KCl and the following: 20 mM imidazole; 20 mM imidazole, 5 mM ATP; or 20 mM imidazole, 5 mM ATP, and 5 mM MgCl2, all at pH 7.0. Skeletal and cardiac myosin filaments formed in imidazole buffer alone were bipolar, tapered at both ends, and about 1.6 micron long. Removal of LC2 resulted in the formation of shorter thick filaments (1.2 micron long). This effect could be reversed by reassociation with LC2. Inclusion of ATP in the buffer disrupted the filament structure, resulting in irregular, short filaments (less than 0.6 micron); addition of both ATP and MgCl2 largely reversed the effects of ATP alone. In cardiac myosin filaments, the bare zone diameter increased from 16 nm as measured in control and LC2-recombined samples to 20 nm in LC2-deficient myosin assemblies. These results implicate LC2 in an active role in controlling synthetic thick filament length in both skeletal and cardiac muscles.

Use of 18O-labelled leucine and phenylalanine to measure protein turnover in muscle cell cultures and possible futile cycling during aminoacylation.

Amino acids labelled with 18(O) on both carboxy oxygen atoms have the potential for use as non-recyclable tracers to measure protein turnover. During protein synthesis one of the labelled oxygen atoms is removed, and thus release of the mono-labelled amino acid could be used to determine proteolysis. Primary cultures of embryonic-chick skeletal-muscle cells were used to test the use of 18(O2)-labelled Leu to measure proteolysis. For 9-day cultures, prelabelled on days 2-8 with medium containing one-half the Leu as [18O2]Leu and one-half as [2H3]Leu, release of [18(O)]Leu was less than 50% that of [2H]Leu over 24 h, suggesting a loss of the 18O label by a mechanism other than protein synthesis. Medium containing [18(O2)]Leu, [2H3]Leu, [18O2]Phe and [13C]Phe was then incubated with 9-day cultures to compare the rate of loss of the 18(O)-label from Leu and Phe with the rate of uptake of the non-carboxy-oxygen-labelled amino acids. Results for Leu demonstrated an 81% loss of the 18(O) label compared with a 33% decrease in [2H]Leu over 12 h. Loss of the 18(O) label was four times as great for Leu as for Phe. Loss of the 18(O) label was not decreased by addition of cycloheximide or by addition of a 3-fold excess of Ile, Val and Tyr; thus the loss of label was not due to protein synthesis alone or to misbinding to incorrect tRNAs. Infusion of the isotopes into pigs showed that the 18(O) label of Leu was not lost during transamination to alpha-ketoisocaproate (alpha-oxoisohexanoate). The most probable explanation is that the 18(O) label is lost as a result of the enzymic deacylation of tRNA, that this process is substantially faster for Leu than for Phe, and that this represents a potentially costly futile cycle for Leu.

A bent monomeric conformation of myosin from smooth muscle.

Smooth muscle myosin filaments formed in 0.15 M KCl are depolymerized by MgATP to a 10S component, rather than to the 6S component typical of myosin monomer in high salt concentrations. This 10S species is also monomeric as determined by sedimentation equilibrium and calculated from the diffusion and sedimentation coefficients. The conformation of 10S myosin is, however, very different from that of 6S myosin, which has a flexible but extended rod. The Stokes radius and the viscosity of 10S myosin are less than those of 6S myosin, consistent with a structure in which the rod is bent. Electron microscopy of rotary-shadowed preparations confirmed that the light meromyosin region of the rod is bent back on subfragment 2, that region of the rod adjacent to the two globular heads. MgATP and dephosphorylation of the 20,000 molecular weight light chain increase the amount of 10S myosin present in 0.15 M KCl; addition of salt converts 10S myosin back to the typical 6S conformation. We conclude that smooth muscle myosin preferentially forms a bent or folded conformation instead of the extended shape usually associated with skeletal muscle myosin, provided that the salt concentration is kept sufficiently low.

ADP-ribosylation of the intermediate filament protein desmin and inhibition of desmin assembly in vitro by muscle ADP-ribosyltransferase.

Arginine-specific mono(ADP-ribosyl)transferase purified from rabbit skeletal muscle catalyzes stoichiometric ADP-ribosylation of the intermediate filament protein, desmin. In contrast, cholera toxin catalyzes a much lower level of ADP-ribosylation of desmin. Modification results in potent inhibition of desmin's ability to assemble into filaments. Phosphorylation of desmin by the catalytic subunit of cAMP dependent protein kinase is also inhibited by ADP-ribosylation. ADP-ribosylation site(s) are located within the N-terminal head domain of desmin.

Substructure of cytoplasmic dense bodies and changes in distribution of desmin and alpha-actinin in developing smooth muscle cells.

The substructure of assembling cytoplasmic dense bodies (CDBs) and changes in the distribution of desmin and alpha-actinin during development of smooth muscle were studied in gizzard samples from 10- and 16-day embryos and from 1- and 7-day post-hatch chickens. CDBs in these cells lack the density of CDBs in mature or adult smooth muscle cells and, thus, allow observations of the changes inside CDBs. The random filament orientation seen in younger embryonic cells is first modified to include relatively small patches of IFs that are somewhat straighter and are approaching a side-by-side arrangement. As development proceeds, the IFs in these arrays become straighter, are parallel over longer lengths of the IFs and later acquire the density characteristic of mature CDBs. Anti-desmin labeling in embryonic 10- and 16-day cells showed that desmin intermediate filaments (IFs) were located in the myofilament compartment but were concentrated in or near assembling CDBs. Anti-desmin labeling shifted to the perimeter of CDBs after hatching. Cross sections, longitudinal sections, and stereo pairs all show that IF profiles are present inside unlabeled assembling CDBs. Anti-alpha-actinin labeling was directly on CDBs and was often associated with the cross-connecting filaments (CCFs) (average diameter of 2-3nm) inside CDBs. We propose, based on these data, that desmin IFs, alpha-actinin-containing CCFs, and actin filaments are the principal components of the substructure of assembling CDBs. We also present a proposed model for CDB assembly.

Synemin may function to directly link muscle cell intermediate filaments to both myofibrillar Z-lines and costameres.

Synemin is a large intermediate filament (IF) protein that has been identified in all types of muscle cells in association with desmin- and/or vimentin-containing IFs. Our previous studies (Bellin, R. M., Sernett, S. W., Becker, B., Ip, W., Huiatt, T. W., and Robson, R. M. (1999) J. Biol. Chem. 274, 29493-29499) demonstrated that synemin forms heteropolymeric IFs with major IF proteins and contains a binding site for the myofibrillar Z-line protein alpha-actinin. By utilizing blot overlay assays, we show herein that synemin also interacts with the costameric protein vinculin. Furthermore, extensive assays utilizing the Gal4 yeast two-hybrid system demonstrate interactions of synemin with desmin and vimentin and additionally define more precisely the protein subdomains involved in the synemin/alpha-actinin and synemin/vinculin interactions. The C-terminal approximately 300-amino acid region of synemin binds to the N-terminal head and central rod domains of alpha-actinin and the approximately 150-amino acid C-terminal tail of vinculin. Overall, these interactions indicate that synemin may anchor IFs to myofibrillar Z-lines via interactions with alpha-actinin and to costameres at the sarcolemma via interactions with vinculin and/or alpha-actinin. These linkages would enable the IFs to directly link all cellular myofibrils and to anchor the peripheral layer of myofibrils to the costameres.

Paraneoplastic IgG striational autoantibodies produced by clonal thymic B cells and in serum of patients with myasthenia gravis and thymoma react with titin.

Autoantibodies with heterogeneous specificities for contractile elements of striated muscle are found in 80 to 90% of patients who have myasthenia gravis (MG) with thymoma. The stimulus for production of these paraneoplastic striational autoantibodies (StrAb) is unknown. One approach to understanding their association with MG and thymoma is to define the antigenic specificities of monoclonal StrAb secreted by B cell lines established from patients who have MG and thymoma. Here, we report the isolation from a single patient of two independent thymic B cell clones secreting StrAb of IgG isotype. In immunoblots, both StrAb bound monospecifically to an antigen of human skeletal muscle that comigrated with the high molecular weight protein, titin. The pattern of immunofluorescence staining yielded by both antibodies in cultured human muscle cells was similar to that produced by rabbit polyclonal anti-titin antibodies. Each monoclonal antibody bound to a different region of the sarcomere in stretched myofibrils; these corresponded to sites previously reported to bind murine anti-titin monoclonal antibodies. The pattern of sarcomere immunostaining produced by combining the two human monoclonal antibodies was indistinguishable from that produced by serum IgG from the patient whose thymus yielded the B cell clones. Thus, the monoclonal antibodies appear to identify two epitopes of titin that are recognized by IgG StrAb in serum. The finding that IgG anti-titin autoantibodies are restricted to serum of MG patients who have thymoma suggests that titin is a major specificity of IgG StrAb. Our additional finding that anti-titin IgG binds to striated elements in medullary myoid cells of the human thymus supports the hypothesis that StrAb represent an intrathymic B cell immune response that is initiated by autoantigens that are rendered immunogenic for helper T cells in the course of noeplastic transformation to thymoma.

Target proteins for arginine-specific mono(ADP-ribosyl) transferase in membrane fractions from chick skeletal muscle cells.

In a previous study, we found that a specific inhibitor of cellular arginine-specific mono(ADP-ribosyl) transferase, meta-iodobenzylguanidine (MIBG), reversibly inhibited both proliferation and differentiation of cultured embryonic chick primary muscle myoblasts. In addition, we observed that arginine-specific ADP-ribosyltransferase activity increased with muscle-cell differentiation in cultures. Therefore, muscle-cell cultures, especially the 96-h myotube cultures that contain the highest levels of ADP-ribosyltransferase, were used as a working system to determine the cellular protein substrates for arginine-specific ADP-ribosyltransferase. When membrane fractions extracted from 96-h chick myotubes were incubated with [32P]NAD at 30 degrees C for 30 min, only a few proteins were labeled. The labeling of two proteins of 36 and 56 kDa was inhibited by the presence of an arginine-specific mono(ADP-ribosyl) transferase inhibitor, MIBG, and by novobiocin. To prove that these proteins are indeed the targets for arginine-specific mono(ADP-ribosyl)ation, active recombinant muscle ADP-ribosyltransferase was incubated with membrane proteins under the same conditions. ADP-ribosylation of these two membrane proteins, as seen in the endogenous reactions, was also catalyzed by the added muscle transferase and was also inhibited by MIBG and novobiocin. By using antibody specific for desmin for immunoprecipitation and immunoblot analysis, we found that a 56-kDa protein associated with the membrane of myotubes is desmin. Our results showed that incorporation of isotope into this protein band from [32P]NAD is due to ADP-ribosylation of desmin.

Determination of the critical concentration required for desmin assembly.

The critical concentration required for filament assembly in vitro from highly purified desmin was determined by both turbidity and centrifugation assays. Assembly was done in the presence of 2 mM-Ca2+, 2 mM-Mg2+ or 150 mM-Na+ at 2, 22 and 37 degrees C. Similar values for critical concentration were obtained by both assays. As temperature increased, critical concentration decreased for each cation. The critical concentration was lowest in the presence of Ca2+ at 2, 22 and 37 degrees C, but was highest in the presence of 150 mM-Na+ at 2 degrees C. Negative staining showed that supernatants from the centrifugation assays contained protofilaments, protofibrils and short particles (less than 300 nm), but pellets contained long filaments (greater than 1 micron) with an average diameter of 10 nm. As the temperature increased, both the average diameter and average length of particles in the supernatant increased. Thermodynamic analysis indicated that hydrophobic interactions were dominant during desmin assembly, but that ionic interactions might also be involved. Our results demonstrated that the specific cation and temperature and temperature-cation interactions all are important in assembly of desmin intermediate filaments.

Assembly of contractile and cytoskeletal elements in developing smooth muscle cells.

Specific developmental changes in smooth muscle were studied in gizzards obtained from 6-, 8-, 10-, 12-, 14-, 16-, 18-, and 20-day chick embryos and from 1- and 7-day posthatch chicks. Myoblasts were actively replicating in tissue from 6-day embryos. Cytoplasmic dense bodies (CDBs) first appeared at Embryonic Day 8 (E8) and were recognized as patches of increased electron density that consisted of actin filaments (AFs), intermediate filaments (IFs), and cross-connecting filaments (CCFs). Although the assembly of CDBs was not synchronized within a cell, the number, size, and electron density of CDBs increased as age increased. Membrane-associated dense bodies (MADBs) also could be recognized at E8. The number and size of MADBs increased as age increased, especially after E16. Filaments with the diameter of thick filaments first appeared at E12. Smooth muscle cells were able to divide as late as E20. The axial intermediate filament bundle (IFB) could first be identified in 1-day posthatch cells and became larger and more prominent in 7-day posthatch cells. Immunogold labeling of 1- and 7-day posthatch cells with anti-desmin showed that the IFB contained desmin IFs. The developmental events during this 23-day period were classified into seven stages, based primarily on the appearance and the growth of contractile and cytoskeletal elements. These stages are myoblast proliferation, dense body appearance, thick filament appearance, dense body growth, muscle cell replication, IFB appearance, and appearance of adult type cells. Smooth muscle cells in each stage express similar developmental characteristics. The mechanism of assembly of myofilaments and cytoskeletal elements in smooth muscle in vivo indicates that myofilaments (AFs and thick filaments) and filament attachment sites (CDBs and MADBs) are assembled before the axial IFB, a major cytoskeletal element.

Synemin contains the rod domain of intermediate filaments.

We have screened a lambda gt11 cDNA library from chicken gizzard muscle with polyclonal antiserum to avian synemin. Immunopositive clones were characterized and sequenced. Computer searches identified primarily intermediate filament (IF) proteins as being homologous to the synemin clones. Synemin's sequence contained all structural features characteristic of the rod domain of IF proteins: (a) its length (approximately 310 amino acids), (b) the heptad repeat pattern of hydrophobic residues of coiled-coil proteins, (c) subdomain structure of the IF rod, by which helical subdomains (1A, 1B, 2A, and 2B) are separated by three short non-helical linker regions, and (d) several potential intrahelical ion pairs along the sequence. We also confirmed the presence of the IF rod domain at the protein level by immunoblotting a proteolytic digest of synemin by using a monoclonal antibody specific to the rod domain of IFs.

The effects of mono-ADP-ribosylation on desmin assembly-disassembly.

Previous studies have shown that desmin, the muscle-specific intermediate filament protein, is a substrate for the endogenous muscle arginine-specific mono-ADP-ribosyltransferase and that ADP-ribosylation inhibits assembly of desmin into intermediate filaments (Huang et al., Exp. Cell Res. 226, 147-153, 1996). In this paper, the effects of mono-ADP-ribosylation on assembly and disassembly of desmin intermediate filaments were further characterized. First, it was found that ADP-ribosylated desmin does not coassemble with unmodified desmin and has no effect on assembly of unmodified desmin. Second, incubation of assembled desmin filaments with mono-ADP-ribosyltransferase and NAD+ results in disassembly of the filaments. Finally, the structural components of the attached ADP-ribose moiety responsible for altering the assembly of desmin into filaments were investigated by a stepwise cleavage of ADP-ribose with snake venom phosphodiesterase and alkaline phophatase, followed by analysis of assembly. The reactions catalyzed by these two enzymes were established using a desmin peptide as a substrate. Our results show that ribosylated desmin, but not phosphoribosylated desmin, was able to self-assemble into intermediate filaments, suggesting that the presence of a phosphate group is needed to alter desmin's assembly ability.

Characterization of ADP-ribosylation sites on desmin and restoration of desmin intermediate filament assembly by de-ADP-ribosylation .

Desmin is an intermediate filament protein that can be ADP-ribosylated by arginine-specific mono(ADP-ribosyl) transferase. Stoichiometric modification of desmin by the transferase causes inhibition of assembly of desmin into 10-nm intermediate filaments (Huang et al., 1993, Biochem. Biophys. Res. Commun. 197, 570-577). In this work, the sites of modification that can affect disassembly have been identified. ADP-ribosylated desmin (1.2 mol ADP-ribose/mol desmin) was digested with lysyl endopeptidase followed by trypsin. Two ADP-ribosylated peptides were obtained, sequenced by Edman degradation, and analyzed by the use of matrix-assisted laser desorption/ionization mass spectrometry. Arginines 48 and 68 of desmin's head domain were shown to be sites of modification, with arginine 48 the major ADP-ribosylation site. ADP-ribosylated desmin (4 mol ADP-ribose/mol desmin) was treated with ADP-ribosylarginine hydrolase. Removal of more than three ADP-ribose groups results in partial restoration of desmin's ability to form intermediate filaments. It is necessary to remove all ADP-ribose groups from desmin to restore its complete ability to form intermediate filaments. The fact that the effect of ADP-ribosylation on the filament-forming properties of desmin is fully reversible suggests that ADP-ribosylation alone is responsible for the changes noted in desmin.